On a March afternoon in Arizona’s Coconino National Forest, artificial intelligence (AI) detected early signs of smoke via a camera feed, prompting human analysts to confirm the observation and alert the state’s forestry service and Arizona Public Service, the largest electric utility in the state. The AI system had identified what became known as the Diamond Fire, allowing firefighters to respond rapidly and contain the blaze before it grew beyond seven acres.

As record-breaking heat and diminished snowpack contribute to heightened wildfire risks, states across the western United States are increasingly turning to AI technology to enhance early wildfire detection. Utilities such as Arizona Public Service have deployed nearly 40 AI-enabled smoke-detection cameras, aiming to expand to 71 by summer’s end. Similarly, Xcel Energy in Colorado has installed 126 cameras and plans to cover seven of the eight states it serves with this technology by the end of the year. California operates ALERTCalifornia, a network of approximately 1,240 AI-equipped cameras monitoring high-risk areas.

Officials emphasize that earlier detection through AI helps fire management teams deploy resources promptly, minimizing fire growth. John Truett, fire management officer for the Arizona Department of Forestry and Fire Management, noted that rapid alerts enable the launch of aircraft and personnel to contain fires at their earliest stages. Likewise, Brent Pascua, battalion chief for the California Department of Forestry and Fire Protection (Cal Fire), highlighted that AI monitoring often identifies fires before 911 calls are made, particularly in remote or sparsely populated areas.

The underlying AI technology typically combines high-definition camera feeds with satellite data and continuous monitoring. Pano AI, a company specializing in this field since 2020, reported detecting 725 wildfires in the U.S. last year alone. Its co-founder Arvind Satyam credited the technology with providing “visual intelligence” that can prevent fires from expanding into hundreds or thousands of acres. Arizona Public Service representatives say their AI system notifies them approximately 45 minutes earlier than the first public emergency call.

Despite its advantages, the technology faces challenges. Cost is a significant barrier, with Pano AI charging about $50,000 annually per camera, which includes fire risk analysis and access to a 24/7 intelligence center. False positives also demand attention and resources, potentially straining firefighting capabilities. Moreover, AI detection does not replace human decision-making critical to determining appropriate response measures, including deployment strategies and evacuation planning. Patrick Roberts, a senior researcher with the nonprofit RAND Corporation, emphasized that human expertise remains essential to interpreting AI alerts and deciding subsequent actions.

The effectiveness of AI in wildfire detection varies by context. In densely populated areas, rapid human reporting often suffices, while extreme weather conditions—such as high winds—can limit the technology’s utility by causing fires to shift unpredictably. Pascua noted that AI complements rather than replaces human judgment, providing real-time information that aids tactical decisions during fire suppression efforts.

Beyond detection, researchers are exploring additional AI applications. For example, teams led by Professor Chaowei “Phil” Yang at George Mason University, collaborating with California State University, Los Angeles, the city of Los Angeles, and NASA’s Jet Propulsion Laboratory, are developing AI systems to forecast fire spread and anticipate communities most impacted by smoke pollution. Such forecasting tools aim to support quick and informed responses, including evacuations, school closures, and air quality warnings, with plans to operationalize the technology within three years.

Overall, experts agree that AI’s role in wildfire management is expanding and evolving. As technologies improve and integrate further with human decision support, they are expected to become a standard element in wildfire detection and response efforts across the United States and beyond.